Comparative performance and profitability of fungicides for managing soybean white mold: a network meta-analysis

• Ano de defesa: 2018
• Autor(a) principal: Barro, Jhonatan Paulo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Sclerotinia sclerotiorum; Sclerotinia sclerotiorum - Controle biológico; Podridão-branca-da-haste; Estatística matemática; Fitopatologia
• Link de acesso: https://locus.ufv.br//handle/123456789/27378

White mold, caused by the fungus Sclerotinia sclerotiorum and also known as Sclerotinia stem rot, is a damaging disease of worldwide importance to several crops including soybean. Fungicides are recommended for the control of white mold in Brazil but the best options and the economic benefits have not been fully explored. The data were obtained from a national cooperative trial network conducted from 2008 to 2017 across 23 locations in Brazil, totaling 72 trials. The five fungicide treatments evaluated in at least 20 trials during four years, and applied twice (flowering - R1 and 10 days later), were: dimoxistrobin+boscalid (DIMO+BOSC), fluazinam (FLUZ), fluopyram (FLUO), procymidone (PROC) and carbendazim+procymidone (CARB+PROC). The sixth treatment was the benzimidazole thiophanate-methyl (TIOF) applied four times starting at R1 stage and every 10-day interval, which was available in 43 trials conducted during seven growing seasons. A network meta-analytic model was fitted to the log of the means of white mold incidence (%) and sclerotia mass (g/ha) data and to the non-transformed mean yield (kg/ha) for each treatment, including the control. The estimated percent reduction in disease incidence relative to the control ranged from 55.2% (TIOF) to 82.8% (CARB+PROC); the latter not differing from FLUO (80.8%) and DIMO+BOSC (80.8%). There was similar percent reduction in sclerotia mass for CARB+PROC (86.4%), DIMO+BOSC (84.9%) and FLUO (83.0%), all performing better than TIOF (53.9%). The mean yield gain ranged from 312 kg/ha (TIOF) to 593 kg/ha (FLUZ); the latter did not differ from DIMO+BOSC (588 kg/ha) and FLUO (551 kg/ha). The model was expanded to include a moderator variable for the incidence level in the check treatment, or the baseline incidence (≤30% or >30%) that explained portion of the heterogeneity. The mean estimates and between-study variance for each baseline disease class were used to calculate the probability of breaking even on fungicide costs. Different scenarios of soybean prices (252 to 404 U$/ton) and fungicide costs (product + application) were created. For TIOF, the range was from 52 to 84 U$/ha (for four applications); for FLUZ, DIMO+BOSC and CARB+PROC the range was from 80 to 112 U$/ha (two applications); and for PROC the range was from 70 to 102 U$/ha (two applications). For the high- disease scenario, probabilities were higher (> 65%) for the more effective/expensive fungicides. For the low-disease scenario, profitability depended on the benefit-cost ratio of fungicides. These results may be useful for decision making in disease management by taking both technical and economic decisions into account.